402 CHEMICAL ENGINEERING wooden tanks, to the proper density for crystallization, then run into crystallizing tanks and allowed to cool and deposit crystals until no more material will crystallize out. In one large plant the acid ferrous-sulphate solution which has been used for removing scale from wire rods is boiled with iron scrap until the acid is completely used up, then run through filters to remove mechanical impurities, and allowed to cool and crystallize in wooden tubs 12 ft. in diameter by 3 ft. deep. When crystallization is complete the mother liquor is drained from the bottom of the tub and returned to the process. The crystals, frequently 18 in. deep, are shoveled out into cars where they drain and dry. They are then ready for use or shipment. In another large plant making copper sulphate and zinc sulphate, the solutions are concentrated in open lead-lined evaporators. The crystallizing tanks in this plant are very large, 5 ft. wide by 2 ft. deep by 60 ft. long. In order therefore that the crystallization may be uniform the batches of concentrated liquors are run into large storage tanks where the liquor is kept hot until enough is accumulated to fill one of the crystallizing tanks. Upon completion of crystallization the mother liquor is drained off and the crystals shoveled out on to an inclined draining platform, which permits the drainage from the crystals to run back into the crystallizing tank. The crystals remain upon the draining platform until dry when they are shovelled into cars and taken to the packing house.1 Crystallization by Agitation of Concentrated Solutions During Cooling.—The above plant produces two grades of zinc sulphate, one the large clear crystals for ordinary commercial use, the other a fine granular crystal such as is found in drug stores and used where superior purity is desired. This type of crystal seems likely to supersede the large irregular crystals of all kinds with which we are most familiar. It has already done so almost entirely in the case of sodium ' bichromate and has been found satisfactory in the case of sodium ferro cyanide (yellow prussiate of soda). These small crystals are produced by artificially cooling the concentrated liquor, at the same time keeping it in constant motion. Under these conditions only the small crystals form. The process is much more rapid than that usually followed, since by the time the liquor is thoroughly cold nearly all the possible crystals have been thrown out, and being so small are much purer than the large crystals. The saving in space, size of equipment and time of production offset the cost of cooling the liquor; and in the best equipped plants using this process it is very nearly continuous, the cold liquor being so stirred that the crystals are kept in suspension and go direct from the crystallizing tank to the filter or centrifugal machine, whence the dried crystals go to the packing room and the mother liquor and wash are returned to process for reconcentra-tion and recooling. The material of which tanks are made has some influence on this point of quick crystallization. The ordinary tanks for copper-sulphate crystallization are of wood, lined with lead. It has been found that the solution will cool much more quickly in metal tanks lined with lead, and that even cutting "windows" in the wood work of the tank, where it is structurally safe to do so, helps quick crystallization. This, though, was not an advantage in former times, as the trade paid fancy prices for large 1 In making double salts, it is probable that better results are obtained if both components are put into solution separately rather than that one should be dissolved in a concentrated solution of the other. The observations may be baseless, but I have always thought for instance, that superior crystals of nickel-ammonium sulphate were produced by bringing both single sulphates into solution separately. —EDITOR.